Performance of Adaptive Beamforming by Using Complex-Valued Neural Network
説明
This paper presents a performance analysis of adaptive beamforming (ABF) by using complex-valued neural network (CVNN). We compare the performance of conventional complex-valued Least Mean Square (CLMS)-based ABF with that of multilayer CVNN’s, using the beamforming results of exact matrix method as a reference. Experiments for multiple beam-pointing and multiple null-steering shows that the CVNN-based ABF outperform the CLMS-based ABF in terms of convergence speed and interferences suppression level. Additionally, the solution of CVNN-based ABF is closer to the exact solution, than the CLMS is.